Patent Grant
8352226
The present invention is related to the optimization of oil and gas field production. More particularly, the present invention is related to the use of a proxy simulator for improving decision making in controlling the operation of oil and gas fields by responding to data as the data is being measured.
Reservoir and production engineers tasked with modeling or managing large oil fields containing hundreds of wells are faced with the reality of only being able to physically evaluate and manage a few individual wells per day. Individual well management may include performing tests to measure the rate of oil, gas, and water coming out of an individual well (from below the surface) over a test period. Other tests may include tests for measuring the pressure above and below the surface as well as the flow of fluid at the surface. As a result of the time needed to manage individual wells in an oil field, production in large oil fields is managed by periodically (e.g., every few months) measuring fluids at collection points tied to multiple wells in an oil field and then allocating the measurements from the collection points back to the individual wells. Data collected from the periodic measurements is analyzed and used to make production decisions including optimizing future production. The collected data, however, may be several months old when it is analyzed and thus is not useful in real time management decisions. In addition to the aforementioned time constraints, multiple analysis tools may be utilized which making it difficult to construct a consistent analysis of a large field. These tools may be multiple physics-based simulators or analytical equations representing oil, gas, and water flow and processing.
In order to improve efficiency in oil field management, sensors have been installed in oil fields in recent years for continuously monitoring temperatures, fluid rates, and pressures. As a result, production engineers have much more data to analyze than was generated from previous periodic measurement methods. However, the increased data makes it difficult for production engineers to react to the data in time to respond to detected issues and make real time production decisions. For example, current methods enable the real time detection of excess water in the fluids produced by a well but do not enable an engineer to quickly respond to this data in order to change valve settings to reduce the amount of water upon detection of the excess water. Further developments in recent years have resulted in the use of computer models for optimizing oil field management and production. In particular, software models have been developed for reservoirs, wells, and gathering system performance in order to manage and optimize production. Typical models used include reservoir simulation, well nodal analysis, and network simulation physics-based or physical models. Currently, the use of physics-based models in managing production is problematic due to the length of time the models take to execute. Moreover, physics-based models must be “tuned” to field-measured production data (pressures, flow rates, temperatures, etc.) for optimizing production. Tuning is accomplished through a process of “history matching,” which is complex, time consuming, and often does not result in producing unique models. For example, the history matching process may take many months for a specialist reservoir or production engineer. Furthermore, current history match algorithms and workflows for assisted or automated history matching are complex and cumbersome. In particular, in order to account for the many possible parameters in a reservoir system that could effect production predictions, many runs of one or more physics-based simulators would need to be executed, which is not practical in the industry.
It is with respect to these and other considerations that the present invention has been made.
Illustrative embodiments of the present invention address these issues and others by providing for real-time oil and gas field production optimization using a proxy simulator. One illustrative embodiment includes a method for establishing a base model of a physical system in one or more physics-based simulators. The physical system may include a reservoir, a well, a pipeline network, and a processing system. The one or more simulators simulate the flow of fluids in the reservoir, well, pipeline network, and a processing system. The method further includes using a decision management system to define control parameters of the physical system for matching with observed data. The control parameters may include a valve setting for regulating the flow of water in a reservoir, well, pipeline network, or processing system. The method further includes defining boundary limits including an extreme level for each of the control parameters of the physical system through an experimental design process, automatically executing the one or more simulators over a set of design parameters to generate a series of outputs, the set of design parameters comprising the control parameters and the outputs representing production predictions, collecting characterization data in a relational database, the characterization data comprising values associated with the set of design parameters and values associated with the outputs from the one or more simulators, fitting relational data comprising a series of inputs, the inputs comprising the values associated with the set of design parameters, to the outputs of the one or more simulators using a proxy model or equation system for the physical system. The proxy model may be a neural network and is used to calculate derivatives with respect to design parameters to determine sensitivities and compute correlations between the design parameters and the outputs of the one or more simulators. The method further includes eliminating the design parameters from the proxy model for which the sensitivities are below a threshold, using an optimizer with the proxy model to determine design parameter value ranges, for the design parameters which were not eliminated from the proxy model, for which outputs from the neural network match observed data, the design parameters which were not eliminated then being designated as selected parameters, placing the selected parameters and their ranges from the proxy model into the decision management system, running the decision management system as a global optimizer to validate the selected parameters in the one or more simulators, and using the proxy model for real time optimization and control decisions with respect to the selected parameters over a future time period.
Other illustrative embodiments of the invention may also be implemented in a computer system or as an article of manufacture such as a computer program product or computer readable media. The computer program product may be a computer storage media readable by a computer system and encoding a computer program of instructions for executing a computer process. The computer program product may also be a propagated signal on a carrier readable by a computing system and encoding a computer program of instructions for executing a computer process.
These and various other features, as well as advantages, which characterize the present invention, will be apparent from a reading of the following detailed description and a review of the associated drawings.
Illustrative embodiments of the present invention provide real-time oil and gas field production optimization using a proxy simulator. Referring now to the drawings, in which like numerals represent like elements, various aspects of the present invention will be described. In particular,
Embodiments of the present invention may be generally employed in the operating environment 100 as shown in
The surface facilities 102 and the wells and subsurface flow devices 104 are in communication with field sensors 106, remote terminal units 108, and field controllers 110, in a manner know to those skilled in the art. The field sensors 106 measure various surface and sub-surface properties of an oilfield (i.e., reservoirs, wells, and pipeline networks) including, but not limited to, oil, gas, and water production rates, water injection, tubing head, and node pressures, valve settings at field, zone, and well levels. In one embodiment of the invention, the field sensors 106 are capable of taking continuous measurements in an oilfield and communicating data in real-time to the remote terminal units 108. It should be appreciated by those skilled in the art that the operating environment 100 may include “smart fields” technology which enables the measurement of data at the surface as well as below the surface in the wells themselves. Smart fields also enable the measurement of individual zones and reservoirs in an oil field. The field controllers 110 receive the data measured from the field sensors 106 and enable field monitoring of the measured data.
The remote terminal units 108 receive measurement data from the field sensors 106 and communicate the measurement data to one or more Supervisory Control and Data Acquisition systems (“SCADAs”) 112. As is known to those skilled in the art, SCADAs are computer systems for gathering and analyzing real time data. The SCADAs 112 communicate received measurement data to a real-time historian database 114. The real-time historian database 114 is in communication with an integrated production drilling and engineering database 116 which is capable of accessing the measurement data.
The integrated production drilling and engineering database 116 is in communication with a dynamic asset model computer system 2. In the various illustrative embodiments of the invention, the computer system 2 executes various program modules for real-time oil and gas field production optimization using a proxy simulator. Generally, program modules include routines, programs, components, data structures, and other types of structures that perform particular tasks or implement particular abstract data types. The program modules include a decision management system (“DMS”) application 24 and a real-time optimization program module 28. The computer system 2 also includes additional program modules which will be described below in the description of
As will be discussed in greater detail below with respect to
Referring now to
It should be understood that the computer system 2 for practicing embodiments of the invention may also be representative of other computer system configurations, including hand-held devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, minicomputers, mainframe computers, and the like. Embodiments of the invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.
The mass storage device 14 is connected to the CPU 5 through a mass storage controller (not shown) connected to the bus 12. The mass storage device 14 and its associated computer-readable media provide non-volatile storage for the computer system 2. Although the description of computer-readable media contained herein refers to a mass storage device, such as a hard disk or CD-ROM drive, it should be appreciated by those skilled in the art that computer-readable media can be any available media that can be accessed by the computer system 2.
By way of example, and not limitation, computer-readable media may comprise computer storage media and communication media. Computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EPROM, EEPROM, flash memory or other solid state memory technology, CD-ROM, digital versatile disks (“DVD”), or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by the computer system 2.
According to various embodiments of the invention, the computer system 2 may operate in a networked environment using logical connections to remote computers, databases, and other devices through the network 18. The computer system 2 may connect to the network 18 through a network interface unit 20 connected to the bus 12. Connections which may be made by the network interface unit 20 may include local area network (“LAN”) or wide area network (“WAN”) connections. LAN and WAN networking environments are commonplace in offices, enterprise-wide computer networks, intranets, and the Internet. It should be appreciated that the network interface unit 20 may also be utilized to connect to other types of networks and remote computer systems. The computer system 2 may also include an input/output controller 22 for receiving and processing input from a number of other devices, including a keyboard, mouse, or electronic stylus (not shown in
As mentioned briefly above, a number of program modules may be stored in the mass storage device 14 of the computer system 2, including an operating system 16 suitable for controlling the operation of a networked personal computer. The mass storage device 14 and RAM 9 may also store one or more program modules. In one embodiment, the DMS application 24 is utilized in conjunction with one or more physics-based simulators 26, real-time optimization module 28, and the physics-based models 30 to optimize production control parameters for real-time use in an oil or gas field. As is known to those skilled in the art, physics-based simulators utilize equations representing physics of fluid flow and chemical conversion. Examples of physics-based simulators include, without limitation, reservoir simulators, pipeline flow simulators, and process simulators (e.g. separation simulators). In the various embodiments of the invention, the control parameters may include, without limitation, valve settings, separation load settings, inlet settings, temperatures, pressure gauge settings, and choke settings, at both well head (surface) and downhole locations. In particular, the DMS application 24 may be utilized for defining sets of control parameters in a physics-based or physical model that are unknown and that may be adjusted to optimize production. As discussed above in the discussion of
Referring now to
The illustrative routine 300 begins at operation 305 where the DMS application 24 executed by the CPU 5, instructs the physics-based simulator 26 to establish a “base” model of a physical system. It should be understood that a “base” model may be a physical or physics-based representation (in software) of a reservoir, a well, a pipeline network, or a processing system (such as a separation processing system) in an oil or gas field based on characteristic data such as reservoir area, number of wells, well path, well tubing radius, well tubing size, tubing length, tubing geometry, temperature gradient, and types of fluids which are received in the physics-based simulator. The physics-based simulator 26, in creating a “base” model, may also receive estimated or uncertain input data such as reservoir reserves. It should be understood that one ore more physics-based simulators 26 may be utilized in the embodiments of the invention.
The routine 300 then continues from operation 305 to operation 310 where the DMS application 24 automatically defines control parameters. As discussed above in the discussion of
Once the control parameters are defined, the routine 300 then continues from operation 310 to operation 315, where the DMS application 24 defines boundary limits for the control parameters. In particular, the DMS application 24 may utilize an experimental design process to define the boundary limits. The boundary limits also include one or more extreme levels (e.g., a maximum, midpoint, or minimum) of values for each control parameter. In one embodiment, the experimental design process utilized by the DMS application 24 may be the well known Orthogonal Array, factorial, or Box-Behnken experimental design processes.
The routine 300 then continues from operation 315 to operation 320 where the DMS application 24 automatically executes the physics-based simulator 26 over the set of control parameters as defined by the boundary limits determined in operation 315. It should be understood that, from this point forward, these parameters will be referred to herein as “design” parameters. In executing the set of design parameters, the physics-based simulator 26 generates a series of outputs which may be used to make a number of production predictions. For instance, the physics-based simulator 26 may generate outputs related to the flow of fluid in a reservoir including, without limitation, pressures, hydrocarbon flow rates, water flow rates, and temperatures which are based on a range of valve setting values defined by the DMS application 24.
The routine 300 then continues from operation 320 to operation 325 where the DMS application 24 collects characterization data in a relational database, such as the integrated production drilling and engineering database 116. The characterization data may include value ranges associated with the design parameters as determined in operation 315 (i.e., the design parameter data) as well as the outputs from the physics-based simulator 26.
The routine 300 then continues from operation 325 to operation 330 where the DMS application 24 utilizes a regression equation to fit the design parameter data (i.e., the relational data of inputs) to the outputs of the physics-based simulator 26 using a proxy model. As used in the foregoing description and the appended claims, a proxy model is a mathematical equation utilized as a proxy for the physics-based models produced by the physics-based simulator 26. Those skilled in the art will appreciate that in the various embodiments of the invention, the proxy model may be a polynomial expansion, a support vector machine, a neural network, or an intelligent agent. An illustrative proxy model which may be utilized in one embodiment of the invention is given by the following equation:
It should be understood that in accordance with an embodiment of the invention, a proxy model may be utilized to simultaneously proxy multiple physics-based simulators that predict flow and chemistry over time.
The routine 300 then continues from operation 330 to operation 335 where the DMS application 24 uses the proxy model to determine sensitivities for the design parameters. As defined herein, “sensitivity” is a derivative of an output of the physics-based simulator 26 with respect to a design parameter within the proxy model. The derivative for each output with respect to each design parameter may be computed on the proxy model equation (shown above). The routine 300 then continues from operation 335 to operation 340 where the DMS application 24 uses the proxy model to compute correlations between the design parameters and the outputs of the physics-based simulator 26.
The routine 300 then continues from operation 340 to operation 345 where the DMS application 24 eliminates design parameters from the proxy model for which the sensitivities are below a threshold. In particular, in accordance with an embodiment of the invention, the DMS application 24 may eliminate a design parameter when the sensitivity or derivative for that design parameter, as determined by the proxy model, is determined to be close to a zero value. Thus, it will be appreciated that one or more of the control parameters which were discussed above in operation 310, may be eliminated as being unimportant or as having a minimal impact. It should be understood that the non-eliminated or important parameters are selected for optimization (i.e., selected parameters) as will be discussed in greater detail in operation 350.
The routine 300 then continues from operation 345 to operation 350 where the DMS application 24 uses the real-time optimization module 28 with the proxy model to determine value ranges for the selected parameters (i.e., the non-eliminated parameters) determined in operation 345. In particular, the real-time optimization module 28 may generate a misfit function representing a squared difference between the outputs from the proxy model and the observed real-time data retrieved from the field sensors 106 and stored in the databases 114 and 116. Illustrative misfit functions for a well which may be utilized in the various embodiments of the invention are given by the following equations:
where wi=weight for well i, wi=weight for time t, sim(i,t)=simulated or normalized value for well i at time t, and his(i,t)=historical or normalized value for well i at time t. It should be understood that the optimized value ranges determined by the real-time optimization module 28 are values for which the misfit function is small (i.e., near zero). It should be further understood that the selected parameters and optimized value ranges are representative of a proxy model which may be executed and validated in the physics-based simulator 26, as will be described in greater detail below.
The routine 300 then continues from operation 350 to operation 355 where the real-time optimization module 28 places the selected parameters (determined in operation 345) and the optimized value ranges (determined in operation 350) back into the DMS application 24 which then executes the physics-based simulator 26 to validate the selected parameters at operation 360. It should be understood that all of the operations discussed above with respect to the DMS application 24 are automated operations on the computer system 2.
The routine 300 then continues from operation 360 to operation 365 where the DMS application 24 uses the proxy model for real time optimization and control. It should be understood that control may include advanced process control decisions or proactive control with respect to the selected parameters over a future time period, depending on a particular field configuration. In particular, in accordance with one embodiment, the DMS application 24 may generate one or more graphical displays showing predicted control parameter settings (e.g., valve settings) for optimizing production in an oil well. An illustrative display is shown in
Referring now to
It will be appreciated that the graphs 410-490 show a prediction of how different valve settings need to be changed over the future time period. For instance, the graph 430 shows that the DMS application 24 has predicted that the valve location “L5” should remain completely open for the initial portion of the future time period and then be completely closed for the latter part of the future time period. It will be appreciated that such a situation may occur based on a prediction that a well is going to produce excess water, thus necessitating that the valve be closed. As another example, the graph 450 shows that the DMS application 24 has predicted that the valve location “L3” should initially remain completely open and then be partially closed for the remainder of the future time period.
Based on the foregoing, it should be appreciated that the various embodiments of the invention include methods, systems, and computer-readable media for real-time oil and gas field production optimization using a proxy simulator. A physics-based simulator in a dynamic asset model computer system is utilized to span the range of possibilities for controllable parameters such as valve settings, separation load settings, inlet settings, temperatures, pressure gauge settings, and choke settings. A decision management application running on the computer system is used to build a proxy model that simulates a physical system (i.e., a reservoir, well, or pipeline network) for making future prediction with respect to the controllable parameters. It will be appreciated that the simulation performed by the proxy model is almost instantaneous, and thus faster than traditional physics-based simulators which are slow and difficult to update. Unlike conventional systems which are reactive, the proxy model described in embodiments of the present invention enable predictions of control parameter settings over a future time period, thereby enabling proactive control.
Although the present invention has been described in connection with various illustrative embodiments, those of ordinary skill in the art will understand that many modifications can be made thereto within the scope of the claims that follow. Accordingly, it is not intended that the scope of the invention in any way be limited by the above description, but instead be determined entirely by reference to the claims that follow.
This patent application claims the benefit of U.S. Provisional Patent Application No. 60/763,971 entitled “Methods, systems, and computer-readable media for real-time oil and gas field production optimization using a proxy simulator,” filed on Jan. 31, 2006 and expressly incorporated herein by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4665398 | Lynch et al. | May 1987 | A |
| 5062068 | Kondo et al. | Oct 1991 | A |
| 5148365 | Dembo | Sep 1992 | A |
| 5182730 | Scherbatskoy | Jan 1993 | A |
| 5315530 | Gerhardt et al. | May 1994 | A |
| 5455780 | Nguyen et al. | Oct 1995 | A |
| 5835883 | Neff et al. | Nov 1998 | A |
| 5841678 | Hasenberg et al. | Nov 1998 | A |
| 5924048 | McCormack et al. | Jul 1999 | A |
| 5992519 | Ramakrishnan et al. | Nov 1999 | A |
| 6088656 | Ramakrishnan et al. | Jul 2000 | A |
| 6128579 | McCormack et al. | Oct 2000 | A |
| 6584368 | Bunkofske et al. | Jun 2003 | B2 |
| 6595294 | Dalsmo et al. | Jul 2003 | B1 |
| 6678569 | Bunkofske et al. | Jan 2004 | B2 |
| 6704696 | Kuramochi et al. | Mar 2004 | B1 |
| 6823296 | Rey-Fabret et al. | Nov 2004 | B2 |
| 6853921 | Thomas et al. | Feb 2005 | B2 |
| 6980940 | Gurpinar et al. | Dec 2005 | B1 |
| 7027968 | Choe et al. | Apr 2006 | B2 |
| 7054752 | Zabalza-Mezghani et al. | May 2006 | B2 |
| 7266456 | De Guzman et al. | Sep 2007 | B2 |
| 7292250 | Sepulveda | Nov 2007 | B2 |
| 7415328 | Volponi | Aug 2008 | B2 |
| 7512543 | Raghuraman et al. | Mar 2009 | B2 |
| 7526463 | Collette et al. | Apr 2009 | B2 |
| 7636613 | Borah et al. | Dec 2009 | B2 |
| 7668707 | Watts et al. | Feb 2010 | B2 |
| 7702409 | Lucas et al. | Apr 2010 | B2 |
| 8025072 | Atkinson et al. | Sep 2011 | B2 |
| 20020013687 | Ortoleva | Jan 2002 | A1 |
| 20020049575 | Jalali et al. | Apr 2002 | A1 |
| 20020177955 | Jalali et al. | Nov 2002 | A1 |
| 20030010208 | Jagger et al. | Jan 2003 | A1 |
| 20030139916 | Choe et al. | Jul 2003 | A1 |
| 20040064425 | Depold et al. | Apr 2004 | A1 |
| 20040104027 | Rossi et al. | Jun 2004 | A1 |
| 20040144565 | Koederitz | Jul 2004 | A1 |
| 20040220790 | Cullick et al. | Nov 2004 | A1 |
| 20040254734 | Zabalza-Mezghani et al. | Dec 2004 | A1 |
| 20050096893 | Feraille et al. | May 2005 | A1 |
| 20060184477 | Hartman et al. | Aug 2006 | A1 |
| 20060224369 | Yang et al. | Oct 2006 | A1 |
| 20070150079 | Blevins et al. | Jun 2007 | A1 |
| 20070156377 | Gurpinar et al. | Jul 2007 | A1 |
| 20070168056 | Shayegi et al. | Jul 2007 | A1 |
| 20070179766 | Cullick et al. | Aug 2007 | A1 |
| 20070179767 | Cullick et al. | Aug 2007 | A1 |
| 20070179768 | Cullick et al. | Aug 2007 | A1 |
| 20080133194 | Klumpen et al. | Jun 2008 | A1 |
| 20080262737 | Thigpen et al. | Oct 2008 | A1 |
| 20090192632 | Yamartino | Jul 2009 | A1 |
| 20100078047 | Labib et al. | Apr 2010 | A1 |
| 20100131257 | Watts et al. | May 2010 | A1 |
| 20110010079 | Shutty et al. | Jan 2011 | A1 |
| Number | Date | Country |
|---|---|---|
| 1 982 046 | Mar 2011 | EP |
| 2398900 | Jan 2004 | GB |
| Number | Date | Country | |
|---|---|---|---|
| 20070192072 A1 | Aug 2007 | US |
| Number | Date | Country | |
|---|---|---|---|
| 60763971 | Jan 2006 | US |
